• R is a statistical programming language (derived from S)
  • Superb data management & graphics capabilities
  • Reproducibility - can keep your scripts to see exactly what was done
  • You can write your own functions
  • Powerful and flexible
  • Runs on all computer platforms

• Well established system of packages and documentation- new enhancements all the time!

• Active development and dedicated community (lots of help)

• Can use a nice GUI front end such as Rstudio

• Can embed your R analyses in dynamic, polished files using R markdown

• FREE

• Resources
  • The R Project Homepage: http://www.r-project.org
  • Quick R Homepage: http://www.statmethods.net
  • Bioconductor: http://www.bioconductor.org
  • An Introduction to R (long!): http://cran.r-project.org/doc/manuals/R-intro.html
  • Google - tutorials, guides, demos, packages and more

• Run R from the command line- type R

• Install a R Integrated Development Environment (IDE)
  • RStudio: http://www.rstudio.com
  • Makes working with R much easier, particularly for a new R user
  • Run on Windows, Mac or Linux OS

• Take a few minutes to familiarize yourself with the R studio environment by locating the following features:
  • The windows clockwise from top left are: the code editor, the workspace and history, the plots and files window, and the R console.
  • In the plots and files window, click on the packages and help tabs to see what they offer.
  • See what types of new files can be made in R studio by clicking the top left icon- open a new R script.
• Now open the file called 'Student_Exercises_for_Workshop_Lectures.Rmd'. This file will serve as your digial notebook for parts of the workshop and contains the other exercises.

• A better way to embed code in R Markdown files
• http://kbroman.org/knitr_knutshell/pages/Rmarkdown.html
• You can insert R chunks into Rmarkdown documents
• Exercise 1.2: Intro to R Markdown Files

• A better way to do this is to assign variables
• Variables are assigned values using the <- operator.
• Variable names must begin with a letter, but other than that, just about anything goes.
• Do keep in mind that R is case sensitive.

x <- 2
x * 3

## [1] 6

y <- x * 3
y - 2

## [1] 4

These do not work

3y <- 3
3*y <- 3

y <- 67
print(y)

## [1] 67

x <- 124
z <- (x*y)^2
print(z)

## [1] 69022864

• Variables and operations can be performed on characters as well
• Note that characters need to be set off by quotation marks to differentiate them from numbers
• The c stands for concatenate
• Note that we are using the same variable names as we did previously, which means that we're overwriting our previous assignment
• A good rule of thumb is to use new names for each variable, and make them short but still descriptive

x <- "I Love"
print (x)

## [1] "I Love"

y <- "Biostatistics"
print (y)

## [1] "Biostatistics"

z <- c(x,y)
print (z)

## [1] "I Love"        "Biostatistics"

• In general R thinks in terms of vectors (a list of characters, factors or numerical values) and it will benefit any R user to try to write programs with that in mind, as it will simplify most things.
• Vectors can be assigned directly using the 'c()' function and then entering the exact values.

n <- c(2,3,4,2,1,2,4,5,10,8,9)
print(n)

##  [1]  2  3  4  2  1  2  4  5 10  8  9

• The vector x is now what is called a list of character values.
• Sometimes we would like to treat the characters as if they were units for subsequent calculations.
• These are called factors, and we can redefine our character variables as factors.
• This might seem a bit strange, but it’s important for statistical analyses where we might want to see the mean or variance for two different treatments.

n_factor <- as.factor(n)
print (n_factor)

z_factor <- as.factor(z)
print (z_factor)

• Note that factor levels are reported alphabetically

str(n)
str(n_factor)
class(n)
class(n_factor)

• We can also determine how R "sees" a variable using str() or class() functions.
• This is a useful check when importing datasets.

• Many functions exist to operate on vectors.

mean(n)
median(n)
var(n)
log(n)
exp(n)
sqrt(n)
sum(n)
length(n)
sample(n, replace = T)

• Notice that the last function (sample) has an argument (replace=T)
• Arguments simply modify or direct the function in some way
• There are many arguments for each function, some of which are defaults

• Getting Help on any function is very easy - just type a question mark and the name of the function.
• There are functions for just about anything within R and it is easy enough to write your own functions if none already exist to do what you want to do.
• In general, function calls have a simple structure: a function name, a set of parentheses and an optional set of parameters to send to the function.
• Help pages exist for all functions that, at a minimum, explain what parameters exist for the function.
  
• Help can be accessed a few ways - try them :

- help(mean)
- ?mean
- example(mean)
- help.search("mean")
- apropos("mean")
- args(mean)

• Here is a way to create your own data sets that are random samples.
• Again, play around with the arguments in the parentheses to see what happens.

• You’ve probably figured out that y from the last example is drawing numbers with equal probability.
• What if you want to draw from a distribution?
• Again, play around with the arguments in the parentheses to see what happens.

x <-rnorm(1000, 0, 100)
hist(x, xlim = c(-500,500))
curve(50000*dnorm(x, 0, 100), xlim = c(-500,500), add=TRUE, col='Red')

- dnorm() generates the probability density, which can be plotted using the curve() function. - Note that is curve is added to the plot using add=TRUE

• So far you've been visualizing just the list of output numbers
• Except for the last example where I snuck in a hist function.
• You can also visualize all of the variables that you've created using the plot function (as well as a number of more sophisticated plotting functions).
• Each of these is called a high level plotting function, which sets the stage
• Low level plotting functions will tweak the plots and make them beautiful

• What do you think that each of the arguments means for the plot function?
• A cool thing about R is that the options for the arguments make sense.
• Try adjusting an argument and see if it works

• On the next slide
• The first line of the lower script tells R that you are going to create a composite figure that has two rows and two columns. Can you tell how?

seq_1 <- seq(0.0, 10.0, by = 0.1)
seq_2 <- seq(10.0, 0.0, by = -0.1)

Say you perform an experiment on two different strains of stickleback fish, one from an ocean population (RS) and one from a freshwater lake (BP) by making them microbe free. Microbes in the gut are known to interact with the gut epithelium in ways that lead to a proper maturation of the immune system.

ADD DATA TABLE

• How should the data set be organized to best analyze it?
• What are the key properties of the variables?
• Why does that matter for learning R?
• Why does that matter for performing statistical analyses?

• Store a copy of data in non-proprietary formats
• Leave an uncorrected file when doing analyses
• Maintain effective metadata about the data
• When you add observations to a database, add rows
• When you add variables to a database, add columns
• A column of data should contain only one data type

• A strength of R is being able to import data from an external source
• Create the same table that you did above in a spreadsheet like Excel
• Export it to comma separated and tab separated text files for importing into R.
• The first will read in a comma-delimited file, whereas the second is a tab-delimited
• In both cases the header and row.names arguments indicate that there is a header row and row label column
• Note that the name of the file by itself will have R look in the CWD, whereas a full path can also be used

YourFile <- read.table('yourfile.csv', header=T, row.names=1, sep=',')
YourFile <- read.csv('yourfile.csv', header=T, row.names=1, sep=',')
YourFile <- read.table('yourfile.txt', header=T, row.names=1, sep='\t')

write.csv(YourFile, "yourfile.csv", quote=F, row.names=T, sep=",")
write.table(YourFile, "yourfile.txt", quote=F, row.names=T, sep="\t")

• Next up - indexing just a subset of the data
• This is a very important idea in R, that you can analyze just a subset of the data.
• This is analyzing only the data in the file you made that has the factor value 'mixed'.

print (YourFile[,2])
print (YourFile$temp)
print (YourFile[2,])
plot (YourFile$temp, YourFile$elevation)

• You can perform operations on particular levels of a factor
• Calculating the mean of the 'mixed' and 'gipps' levels of habitat.
• Note that the first argument is the numerical column vector, and the second is the factor column vector.
• The third is the operation. Reversing the first two does not work (the one below).

tapply(YourFile$temp, YourFile$habitat, mean)
tapply(YourFile$temp, YourFile$habitat, var)

• Hadley Wickham and others have written R packages to modify data

• These packages do many of the same things as base functions in R

• However, they are specifically designed to do them faster and more easily

• Wickham also wrote the package GGPlot2 for elegant graphics creations

• GG stands for ‘Grammar of Graphics’

library(ggplot2)
ggplot(data=diamonds) +
  geom_bar(mapping=aes(x=cut))

Now try this…

ggplot(data=diamonds) +
  geom_bar(mapping=aes(x=cut, colour=cut))

and this…

ggplot(data=diamonds) +
  geom_bar(mapping=aes(x=cut, fill=cut))

and finally this…

ggplot(data=diamonds) +
  geom_bar(mapping=aes(x=cut, fill=clarity), position="dodge")

• With this function you can make a histogram

ggplot(data=diamonds) +
  geom_histogram(mapping=aes(x=carat), binwidth=0.5)

• This allows you to make a frequency polygram

ggplot(data=diamonds) +
  geom_freqpoly(mapping=aes(x=carat), binwidth=0.5)

Boxplots are very useful for visualizing data

ggplot(data=diamonds, mapping=aes(x=cut, y=price)) +
  geom_boxplot()

ggplot(data=mpg, mapping=aes(x=class, y=hwy)) +
  geom_boxplot() +
  coord_flip()

ggplot(data=mpg, mapping=aes(x=reorder(class, hwy, FUN=median), y=hwy)) +
  geom_boxplot() +
  coord_flip()

ggplot(data=diamonds, mapping=aes(x=x, y=y)) +
  geom_point()

ggplot(data=mpg) +
  geom_point(mapping=aes(x=displ, y=hwy)) +
  facet_wrap(~class, nrow=2)

ggplot(data=mpg) +
  geom_point(mapping=aes(x=displ, y=hwy)) +
  facet_grid(drv~cyl)

ggplot(data=mpg) +
  geom_smooth(mapping=aes(x=displ, y=hwy))

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

ggplot(data=mpg) +
  geom_point(mapping=aes(x=displ, y=hwy)) +
  geom_smooth(mapping=aes(x=displ, y=hwy))

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

ggplot(data=mpg, aes(displ, hwy)) +
  geom_point(aes(color=class)) +
  geom_smooth(se=FALSE) +
  labs(
    title = "Fuel efficiency generally decreases with engine size",
    subtitle = "Two seaters (sports cars) are an exception because of their light weight",
    caption = "Data from fueleconomy.gov"
  )

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'